Hi Sergei,

On Tue, Oct 27, 2015, at 20:01, Sergei Shtylyov wrote:

On 10/27/2015 06:02 PM, Hannes Frederic Sowa wrote:

quoted

CHECKSUM_PARTIAL skbs should never arrive in ip_fragment. If we get one
of those warn about them once and handle them gracefully by recalculating
the checksum.

Cc: Eric Dumazet <edumazet@google.com>
Cc: Vlad Yasevich <redacted>
Cc: Benjamin Coddington <redacted>
Cc: Tom Herbert <redacted>
Signed-off-by: Hannes Frederic Sowa <redacted>
---
  net/ipv4/ip_output.c | 8 +++++---
  1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/net/ipv4/ip_output.c b/net/ipv4/ip_output.c
index 0b02417..3f94a3b 100644
--- a/net/ipv4/ip_output.c
+++ b/net/ipv4/ip_output.c

@@ -533,6 +533,11 @@ int ip_do_fragment(struct net *net, struct sock *sk, struct sk_buff *skb,

  	dev = rt->dst.dev;

+	/* for offloaded checksums cleanup checksum before fragmentation */
+	if (WARN_ON_ONCE(skb->ip_summed == CHECKSUM_PARTIAL) &&
+	    (err = skb_checksum_help(skb)))

    scripts/checkpatch.pl shou;d have complained about using = in the
    *if* 
expression.

I know and I ignored it deliberately because I found it nicer this way.
I made sure gcc does not complain by using extra braces around the
assignment.

Bye,
Hannes